is it possible to have 2.5% fee, boost hasrate and support [XMG] Magi ? it ll be great thanks in advance br 1. As always, I'm ready to reduce dev fee if someone who is not too lazy creates a good AMD miner to make some competition. 2. XMG requires some efforts, I will do it when I have time, right now I'm really busy with other miners.
|
|
|
v9.1 is available:
- Improved pool-related routines: less disconnects, more hashrate on some pools. - Fixed issue with some old 6xxx/5xxx cards ("lowDifficultShare" error).
Thx works fine on my 5970 510+h/s Cat 13.12 I am sorry to disappoint but the problem is fixed at only 10%. I have about 10 cores from 58** serie and only 1 and sometimes 2 cores dont make problems. And the problem is not in the cards or in the rigs. Everything is on stock settings and otherwise working just fine. But with v9 and v9.1 (i cant remember what happens with v8) it still gives a lot of server responce errors. The funny part is that they are all pool dependent. Least errors I get with cryptoescrow at 5000 diff. I use 14.6 diff RCs. Should I update the driver to 14.9? Which drivers did u test with? P.S. With all my other AMD cards I have absolutely no problems and I am very happy with the hash increase. 3 of the rings are mixed 5850 cards with R9 serie cards. R9 cards work super. But when it comes to accepting the share from the 5850 card - the usual server responce error appears and it says the this share is with too low difficulty... Make sure that you use "-a 3" for old cards like 6xxx and 5xxx. If it does not help, PM me the log file.
|
|
|
v9.1 is available:
- Improved pool-related routines: less disconnects, more hashrate on some pools. - Fixed issue with some old 6xxx/5xxx cards ("lowDifficultShare" error).
9.1 no linux version? I'm working on it, will be available next week.
|
|
|
v9.1 is available:
- Improved pool-related routines: less disconnects, more hashrate on some pools. - Fixed issue with some old 6xxx/5xxx cards ("lowDifficultShare" error).
|
|
|
I've seen the dev fee was decreased in GPU miner to 2.5%. Will you consider decreasing also the CPU miner fee any time soon?
Yes, I will release an update in 2-3 days, it will also include some other improvements.
|
|
|
v 9.0 have many mistake below..
buf: {"jsonrpc":"2.0","method":"job","params":
And I still did not get any logs from you... This text is not error.
|
|
|
V 9.0 gives me about 13.5% better hashrate across all of my rigs. At least that is what miner shows, but I've left it mining for some time now and pool shows only half of hashrate. Also, one of my rigs reports "SOMETHING WRONG WITH DEVFEE CONNECTION" form time to time. I didn't have same issue wih V 6.1. Anyone else has issues like that?
PM me the log file, I'll check it. I too see today half of hashrate on pool.. version 9.0, catalyst 13.12. It can be pool-related problem. But you can PM me the log file, I'll check it.
|
|
|
Have any 6xx and 5xx cards. Minig with your miner, thanks :-) Miner v8.0 working well. Trying 9.0, see that: Error in server response : {"id":1,"jsonrpc":"2.0","error":{"code":-1,"message":"Lowdifficultyshare"}} Catalyst 13.12. Why? Turning no help me. Returning to 8.0. Slowly, but good:-)))
P.S. With 270x-280x cards 9.0 all ok!
Same here with 14.6 and 14.7 different RCs. I have 4x 5850 and 2x 5990 in different rigs. Every card displays the same error in server response. I was very happy at first because the 5850 made 212 hs at stock with v9.0. But then I realised sum ting was wong... It will be fixed soon, I'm working on it. PM me if you want to test updated version or wait until I release it in public.
|
|
|
V 9.0 gives me about 13.5% better hashrate across all of my rigs. At least that is what miner shows, but I've left it mining for some time now and pool shows only half of hashrate. Also, one of my rigs reports "SOMETHING WRONG WITH DEVFEE CONNECTION" form time to time. I didn't have same issue wih V 6.1. Anyone else has issues like that?
PM me the log file, I'll check it.
|
|
|
2GB, Pitcairn or something like that
1. Try to set "GPU_MAX_ALLOC_PERCENT" environment variable as "60" or "80". Also try to remove this variable at all. 2. Try to test if it still fails if you install more RAM (if you can).
|
|
|
My rigs are with R9 270. Miner v9.0 and v8.0 keep crashing. It work 1-2 hours before crash. After restart it works another 1-2 hours. Version 7.0 is working without any problem.
1. How many GPUs in rig? 2. How much RAM installed? 6 GPU on every rig and 4GB of RAM, virtual memory is set on 17000MB. I can set it 20k-30k Cards R9 270 2GB ? Or R9 270 1GB?
|
|
|
My rigs are with R9 270. Miner v9.0 and v8.0 keep crashing. It work 1-2 hours before crash. After restart it works another 1-2 hours. Version 7.0 is working without any problem.
1. How many GPUs in rig? 2. How much RAM installed?
|
|
|
Hi Claymore, will there be a new version for Linux? The latest available is 7.0, which crashes for me after a day of hashing, sometimes it lasts 2 days, sometimes it crashes an hour after the start.. and I've got a 280x with Elpida chips in my linux box, so I suppose the newer version should also be faster than the current 7.0... not to mention the fee difference. Yes I will release v9.0 for Linux in a few days.
|
|
|
I think you right, it's 13.12. Pasting ocl's from 14.7 in miner folder still works ?
I released two builds of v9.0: for Catalyst 13.12 and for Catalyst 14.x. Just download miner for Catalyst 13.12 if you use Catalyst 13.12, it is about 1% faster than miner for Catalyst 14.x.
|
|
|
With 8.0 it works ok. Probably you are trying to start miner for Catalyst 14.7 on the system with Catalyst 13.x installed. Use appropriate miner version. Also make sure that you updated all files, not only EXE file.
|
|
|
Have any 6xx and 5xx cards. Minig with your miner, thanks :-) Miner v8.0 working well. Trying 9.0, see that: Error in server response : {"id":1,"jsonrpc":"2.0","error":{"code":-1,"message":"Lowdifficultyshare"}} Catalyst 13.12. Why? Turning no help me. Returning to 8.0. Slowly, but good:-)))
P.S. With 270x-280x cards 9.0 all ok!
Send me first 20-30 lines of the log so I can see what card you use and "-a" and "-h" values.
|
|
|
I tried all mods for 270x(Elpida) and -a4 works best so far. -h is 512 by default, i did not change it.
I have 270 and 270X, both Elpida. For 270X I have 445h/s for "-a 1" and 436h/s for "-a 4". Well, the difference is small, so may be in some cases "-a 4" can be better. Don't forget to use "m" key to see exact immediate hashrate. Main hashrate that miner shows can be less because miner calculates it as TotalHashes/TotalTime, hashes are not growing when miner is connecting to pool or when it is disconnected, but TotalTime is always growing. Therefore main hashrate value becomes less if pool is not working properly. But when you press "m" key you see "raw" hashrate based on one GPU round only, it is useful to compare hashrates for different "-a" and "-h" values.
|
|
|
WOW . This new Claymore blows away russian miner. On clay 9.0 1080/1250 270x 444hs -a4, same settings on russian miner -m 2 404hs. Can't wait to try 280x
"-a 1" is faster for 270/270X than "-a 4". For 290X/290/280X/270X/270 you can leave default parameters, i.e. do not specify "-a" and "-h", miner will apply best settings automatically. For other cards, guide how to get maximal speed:1. Try "-a 1", "-a 3", "-a 4" and select best mode, press "m" key to see immediate hashrate. 2. Remember "HashCnt" value when miner starts, then specify "-h" parameter with this value, remember hashrate and change "-h" parameter step by step by 32 in both directions (less and more). Select best value based on immediate hashrate.
|
|
|
any news for linux multi-gpu fixes?
1. Multi-GPU: I tested miner on Ubuntu and Cent, on a system with 3 GPUs and I cannot duplicate this issue. So it must be something related to Catalyst version, or to Linux version that you use. 2. "Input OpenCL binary is not for the target!" error is not related to memory allocation error. It means that Catalyst cannot compile kernels, I can duplicate this issue only if I try to start miner for Catalyst 13.12 on the system with Catalyst 14.x or vice versa.
|
|
|
|